在《pytest封神之路第三步 精通fixture》和《pytest封神之路第四步 内置和自定义marker》两篇文章中，都提到了pytest参数化。那么本文就趁着热乎，赶紧聊一聊pytest的参数化是怎么玩的。
@pytest.mark.parametrize
@pytest.mark.parametrize("test_input,expected", [("3+5", 8), ("2+4", 6), ("6*9", 42)])deftest_eval(test_input, expected):asserteval(test_input) == expected
可以自定义变量，test_input对应的值是"3+5" "2+4" "6*9"，expected对应的值是8 6 42，多个变量用tuple，多个tuple用list
参数化的变量是引用而非复制，意味着如果值是list或dict，改变值会影响后续的test
重叠产生笛卡尔积
importpytest@pytest.mark.parametrize("x", [0, 1])@pytest.mark.parametrize("y", [2, 3])deftest_foo(x, y):pass
@pytest.fixture()
@pytest.fixture(scope="module", params=["smtp.gmail.com", "mail.python.org"])defsmtp_connection(request):smtp_connection = smtplib.SMTP(request.param,587, timeout=5)
只能使用request.param来引用
参数化生成的test带有ID，可以使用-k来筛选执行。默认是根据函数名[参数名]来的，可以使用ids来定义
// list@pytest.fixture(params=[0, 1], ids=["spam", "ham"])// function@pytest.fixture(params=[0, 1], ids=idfn)
使用--collect-only命令行参数可以看到生成的IDs。
参数添加marker
我们知道了参数化后会生成多个tests，如果有些test需要marker，可以用pytest.param来添加
marker方式
# content of test_expectation.pyimportpytest@pytest.mark.parametrize("test_input,expected", [("3+5",8), ("2+4",6), pytest.param("6*9",42, marks=pytest.mark.xfail)],)deftest_eval(test_input, expected):asserteval(test_input) == expected
fixture方式
# content of test_fixture_marks.pyimportpytest@pytest.fixture(params=[0, 1, pytest.param(2, marks=pytest.mark.skip)])defdata_set(request):returnrequest.paramdeftest_data(data_set):pass
pytest_generate_tests
用来自定义参数化方案。使用到了hook，hook的知识我会写在《pytest hook》中，欢迎关注公众号dongfanger获取最新文章。
# content of conf.pydefpytest_generate_tests(metafunc):if"test_input"inmetafunc.fixturenames: metafunc.parametrize("test_input", [0,1])
# content of test.pydeftest(test_input):asserttest_input ==0
定义在conftest.py文件中
metafunc有5个属性，fixturenames，module，config，function，cls
metafunc.parametrize() 用来实现参数化
多个metafunc.parametrize() 的参数名不能重复，否则会报错
参数化误区
在讲示例之前，先简单分享我的菜鸡行为。假设我们现在需要对50个接口测试，验证某一角色的用户访问这些接口会返回403。我的做法是，把接口请求全部参数化了，test函数里面只有断言，伪代码大致如下
defapi():params = []deffunc():returnrequest() params.append(func) ...@pytest.mark.parametrize('req', api())deftest():res = req()assertres.status_code ==403
这样参数化以后，会产生50个tests，如果断言失败了，会单独标记为failed，不影响其他test结果。咋一看还行，但是有个问题，在回归的时候，可能只需要验证其中部分接口，就没有办法灵活的调整，必须全部跑一遍才行。这是一个相对错误的示范，至于正确的应该怎么写，相信每个人心中都有一个答案，能解决问题就是ok的。我想表达的是，参数化要适当，不要滥用，最好只对测试数据做参数化。
实践
本文的重点来了，参数化的语法比较简单，实际应用是关键。这部分通过11个例子，来实践一下。示例覆盖的知识点有点多，建议留大段时间细看。
1.使用hook添加命令行参数--all，"param1"是参数名，带--all参数时是range(5) == [0, 1, 2, 3, 4]，生成5个tests。不带参数时是range(2)。
# content of test_compute.pydeftest_compute(param1):assertparam1 <4
# content of conftest.pydefpytest_addoption(parser):parser.addoption("--all", action="store_true", help="run all combinations")defpytest_generate_tests(metafunc):if"param1"inmetafunc.fixturenames:ifmetafunc.config.getoption("all"): end =5else: end =2metafunc.parametrize("param1", range(end))
2.testdata是测试数据，包括2组。test_timedistance_v0不带ids。test_timedistance_v1带list格式的ids。test_timedistance_v2的ids为函数。test_timedistance_v3使用pytest.param同时定义测试数据和id。
# content of test_time.pyfromdatetimeimportdatetime, timedeltaimportpytesttestdata = [ (datetime(2001,12,12), datetime(2001,12,11), timedelta(1)), (datetime(2001,12,11), datetime(2001,12,12), timedelta(-1)),]@pytest.mark.parametrize("a,b,expected", testdata)deftest_timedistance_v0(a, b, expected):diff = a - bassertdiff == expected@pytest.mark.parametrize("a,b,expected", testdata, ids=["forward", "backward"])deftest_timedistance_v1(a, b, expected):diff = a - bassertdiff == expecteddefidfn(val):ifisinstance(val, (datetime,)):# note this wouldn't show any hours/minutes/secondsreturnval.strftime("%Y%m%d")@pytest.mark.parametrize("a,b,expected", testdata, ids=idfn)deftest_timedistance_v2(a, b, expected):diff = a - bassertdiff == expected@pytest.mark.parametrize("a,b,expected", [ pytest.param( datetime(2001,12,12), datetime(2001,12,11), timedelta(1), id="forward"), pytest.param( datetime(2001,12,11), datetime(2001,12,12), timedelta(-1), id="backward"), ],)deftest_timedistance_v3(a, b, expected):diff = a - bassertdiff == expected
3.兼容unittest的testscenarios
# content of test_scenarios.pydefpytest_generate_tests(metafunc):idlist = [] argvalues = []forscenarioinmetafunc.cls.scenarios: idlist.append(scenario[0]) items = scenario[1].items() argnames = [x[0]forxinitems] argvalues.append([x[1]forxinitems]) metafunc.parametrize(argnames, argvalues, ids=idlist, scope="class")scenario1 = ("basic", {"attribute":"value"})scenario2 = ("advanced", {"attribute":"value2"})classTestSampleWithScenarios:scenarios = [scenario1, scenario2]deftest_demo1(self, attribute):assertisinstance(attribute, str)deftest_demo2(self, attribute):assertisinstance(attribute, str)
4.初始化数据库连接
# content of test_backends.pyimportpytestdeftest_db_initialized(db):# a dummy testifdb.__class__.__name__ =="DB2": pytest.fail("deliberately failing for demo purposes")
# content of conftest.pyimportpytestdefpytest_generate_tests(metafunc):if"db"inmetafunc.fixturenames: metafunc.parametrize("db", ["d1","d2"], indirect=True)classDB1:"one database object"classDB2:"alternative database object"@pytest.fixturedefdb(request):ifrequest.param =="d1":returnDB1()elifrequest.param =="d2":returnDB2()else:raiseValueError("invalid internal test config")
5.如果不加indirect=True，会生成2个test，fixt的值分别是"a"和"b"。如果加了indirect=True，会先执行fixture，fixt的值分别是"aaa"和"bbb"。indirect=True结合fixture可以在生成test前，对参数变量额外处理。
importpytest@pytest.fixturedeffixt(request):returnrequest.param *3@pytest.mark.parametrize("fixt", ["a", "b"], indirect=True)deftest_indirect(fixt):assertlen(fixt) ==3
6.多个参数时，indirect赋值list可以指定某些变量应用fixture，没有指定的保持原值。
# content of test_indirect_list.pyimportpytest@pytest.fixture(scope="function")defx(request):returnrequest.param *3@pytest.fixture(scope="function")defy(request):returnrequest.param *2@pytest.mark.parametrize("x, y", [("a", "b")], indirect=["x"])deftest_indirect(x, y):assertx =="aaa"asserty =="b"
7.兼容unittest参数化
# content of ./test_parametrize.pyimportpytestdefpytest_generate_tests(metafunc):# called once per each test functionfuncarglist = metafunc.cls.params[metafunc.function.__name__] argnames = sorted(funcarglist[0]) metafunc.parametrize( argnames, [[funcargs[name]fornameinargnames]forfuncargsinfuncarglist] )classTestClass:# a map specifying multiple argument sets for a test methodparams = {"test_equals": [dict(a=1, b=2), dict(a=3, b=3)],"test_zerodivision": [dict(a=1, b=0)], }deftest_equals(self, a, b):asserta == bdeftest_zerodivision(self, a, b):withpytest.raises(ZeroDivisionError): a / b
8.在不同python解释器之间测试对象序列化。python1把对象pickle-dump到文件。python2从文件中pickle-load对象。
"""
module containing a parametrized tests testing cross-python
serialization via the pickle module.
"""importshutilimportsubprocessimporttextwrapimportpytestpythonlist = ["python3.5","python3.6","python3.7"]@pytest.fixture(params=pythonlist)defpython1(request, tmpdir):picklefile = tmpdir.join("data.pickle")returnPython(request.param, picklefile)@pytest.fixture(params=pythonlist)defpython2(request, python1):returnPython(request.param, python1.picklefile)classPython:def__init__(self, version, picklefile):self.pythonpath = shutil.which(version)ifnotself.pythonpath: pytest.skip("{!r} not found".format(version)) self.picklefile = picklefiledefdumps(self, obj):dumpfile = self.picklefile.dirpath("dump.py") dumpfile.write( textwrap.dedent(r"""
import pickle
f = open({!r}, 'wb')
s = pickle.dump({!r}, f, protocol=2)
f.close()
""".format( str(self.picklefile), obj ) ) ) subprocess.check_call((self.pythonpath, str(dumpfile)))defload_and_is_true(self, expression):loadfile = self.picklefile.dirpath("load.py") loadfile.write( textwrap.dedent(r"""
import pickle
f = open({!r}, 'rb')
obj = pickle.load(f)
f.close()
res = eval({!r})
if not res:
raise SystemExit(1)
""".format( str(self.picklefile), expression ) ) ) print(loadfile) subprocess.check_call((self.pythonpath, str(loadfile)))@pytest.mark.parametrize("obj", [42, {}, {1: 3}])deftest_basic_objects(python1, python2, obj):python1.dumps(obj) python2.load_and_is_true("obj == {}".format(obj))
9.假设有个API，basemod是原始版本，optmod是优化版本，验证二者结果一致。
# content of conftest.pyimportpytest@pytest.fixture(scope="session")defbasemod(request):returnpytest.importorskip("base")@pytest.fixture(scope="session", params=["opt1", "opt2"])defoptmod(request):returnpytest.importorskip(request.param)
# content of base.pydeffunc1():return1
# content of opt1.pydeffunc1():return1.0001
# content of test_module.pydeftest_func1(basemod, optmod):assertround(basemod.func1(),3) == round(optmod.func1(),3)
10.使用pytest.param添加marker和id。
# content of test_pytest_param_example.pyimportpytest@pytest.mark.parametrize("test_input,expected", [ ("3+5",8), pytest.param("1+7",8, marks=pytest.mark.basic), pytest.param("2+4",6, marks=pytest.mark.basic, id="basic_2+4"), pytest.param("6*9",42, marks=[pytest.mark.basic, pytest.mark.xfail], id="basic_6*9"), ],)deftest_eval(test_input, expected):asserteval(test_input) == expected
11.使用pytest.raises让部分test抛出Error。
fromcontextlibimportcontextmanagerimportpytest//3.7+fromcontextlibimportnullcontextasdoes_not_raise@contextmanagerdefdoes_not_raise():yield@pytest.mark.parametrize("example_input,expectation", [ (3, does_not_raise()), (2, does_not_raise()), (1, does_not_raise()), (0, pytest.raises(ZeroDivisionError)), ],)deftest_division(example_input, expectation):"""Test how much I know division."""withexpectation:assert(6/ example_input)isnotNone
深圳网站建设www.sz886.com
